|
|
||||||||||
|
|
DB development |
|||||||||
|
Identify
all sources of input and decide on desired output
|
Database design begins with the identification of all input sources: forms, phone messages, handwritten data entry, inventory, lab reports & requests. Current forms are invaluable for identifying data. Sometimes design is only a matter of rearranging existing data and creating more meaningful formats. Other times it involves changing or augmenting data -- summarizing, analyzing and performing calculations. Identification of all output (reports, statistics, labels, etc.) to be generated is also important. Steps in the design approach are as follows: | |||||||||
|
Normalization reduces redundancy and aids analysis
|
|
|||||||||
| There are two mechanisms for normalizing data. First, develop and refine the data list or entities. In most cases each entity needs a table of its own. Second, use general rules of thumb to determine when a table needs to be split. If identical data is being entered into two or more records the DB probably needs to be split. Every record should contain data specific to the item, person or event it represents. Avoid storing multiples of a particular item of information in a table. [Top] | ||||||||||